Snake evolution

I often run across the claim that pit vipers represent the apex of snake evolution. Although I've discussed this before, I'd like to say a bit more.

i) Infrared vision is clearly useful to a nocturnal predator. If, however, a particular snake species is diurnal, then that would be a useless refinement. 

ii) Are retractable fangs more advanced than fixed fangs? Certainly retractable fangs are way cool. To my knowledge, extant snake species have at least one of four kinds of fangs:

Vipers have retractable fangs that resemble hypodermic needles. They have to be retractable because they are so long, and located in the front of the mouth. 

The Stiletto has external fangs which enables it to strike sideways or backwards. Is that more or less advanced than the viper? 

Likewise, is the spitting cobra more or less advanced than the viper? 

Then you have elapids, with short, fixed fangs positioned in the front of the mouth. 

Finally, you have rear-fanged snakes (e.g. boomslang). 

iii) Is one design more efficient or advanced than another? I'm no expert, but here are some considerations that come to mind.

Retractable fangs deploy a rapid strike-and-release technique. That means it can envenomate with a glancing strike. Stabbing or scratching the skin. 

That makes it easier to envenomate prey. By contrast, fixed-fanged snakes, and especially rear-fanged snakes, must bite into the prey and hold onto the prey or even chew on the prey to inject venom. 

I imagine it would be harder for a snake with fixed fangs to puncture a flat surface, round surface, or surface with a large circumference, if it can't open its mouth wide enough to puncture that surface.

However, many venomous snakes have a similar diet of small rodents, lizards, amphibians, birds, fish, or even other snakes. So I don't think it makes much difference when that's the quarry. The size and shape of the prey in those cases is conducive to either design.

In fact, there might be types of prey where the strike-and-release technique is disadvantageous. Take the boomslang. That's an arboreal snake that eats birds. Presumably, it would be less effective to let go of the bird, which might fly away or fall to the ground.

Likewise, take sea snakes. Releasing the fish would give it a chance to swim away or be eaten by another opportunistic fish. By the same token, King cobras eat other snakes, including cobras. But the strike-and-release technique would give the prey a chance to get away. 

Conversely, I've read that weasels are on the menu of Timber rattlesnakes. Weasels are predators in their own right. The strike-and-release technique might be beneficial when attacking prey like that, because a weasel could injure the snake if it had to hang onto the weasel until the prey become immobilized. Weasels are very feisty animals which would bite and claw the snake if it had to keep a grip on the prey to inject venom.  

So it isn't clear to me that one design is more advanced than another. They all have tradeoffs. They are all adapted to the nature of the prey. And particular kinds of prey may favor a particular envenomation mechanism. So I don't think that's evidence for macroevolution.